This revision of a competitive renewal proposes to investigate memory deficits in aging and Alzheimer's Disease (AD) using behavioral observations, event-related potentials (ERP) (functional) recording, and magnetic resonance imaging (MRI) (structural) imaging. Memory deficits are a major feature of AD. Semantic memory, the permanent knowledge of objects, facts, and concepts, is impaired early in AD, though knowledge may be lost for only certain items while retained for others. This semantic memory loss may be due to a breakdown in the structure of knowledge itself or may reflect inability to access or retrieve this knowledge. Failure of memory for recently presented information may be due to poor encoding, or poor retrieval. Associations and dissociations between behavioral and ERP responses, and between ERP responses and MRI measures of specific brain regions from normal and impaired subjects will be used to develop a better understanding of specific aspects of memory failures and their neuroanatomical bases. Gender and education-matched groups of healthy young and elderly subjects and age- gender and education-matched healthy subjects and patients with AD will participate in one of two separate studies. The stability of item-specific loss will be assessed in the healthy elderly and AD subjects. In the first study, behavioral and ERP measures of semantic priming to visually presented words and pictures of objects will be obtained. In the second study, behavioral and ERP measures of repetition priming and recognition memory of visually presented words and pictures of objects will be obtained. Healthy elderly and AD subjects will also have a brain MRI scan from which the volume of specific neuroanatomic structures and gray matter in specific cortical regions will be measured and related to electrophysiological measures. Data from these studies will be used to: 1)Establish the reliability of intact and degraded semantic knowledge in AD, the role of knowledge loss vs. access/retrieval difficulties in poor semantic memory, and the role of automatic and strategic processes in access/retrieval of intact and degraded knowledge. 2) Assess whether failures of memory for recently presented information are associated with different ERP patterns of brain activation during encoding of the information and/or at retrieval. 3) Establish brain structure/function associations and dissociations in elderly and AD subjects.